MXPA99012006A - Control of switched reluctance machines - Google Patents

Abstract

Una máquina de reluctancia conmutada polifásica controlada por un sistema de control que utiliza detección de posición sin sensores. El controlador es robusto y confiable y opera en todo el intervalo de velocidad de la máquina. En el modo de corte, son inyectados impulsos de diagnóstico de enlace de flujo predeterminado en una fase desocupada. En el modo de un solo impulso, la predicción de la posición se efectúa utilizando una fase activa. Un método de arranque de la máquina utiliza impulsos de diagnóstico en dos fases para proporcionar un valorúnico para la posición, permitiendo que el dispositivo motor arranque o rearranque después de una torsión completa.

Description

CONTROL OF COMMUTED RELUCTANCE MACHINES DESCRIPTION OF THE INVENTION The present invention relates to the control of switched reluctance machines, particularly those machines which are operated without a sensor to measure the position of the rotor. In general, a reluctance machine is an electrical machine in which the torsion is produced by the tendency of its moving part to move to a position where the reluctance of the magnetic circuit is minimized, that is, where the inductance of the exciting coil is Maximize. In a type of reluctance machine, the circuit is provided to detect the angular position of the rotor and energize the phase coils as a function of the position of the rotor. This is generally known as a switched reluctance machine and can be operated as a motor or a generator. The characteristics of such switched reluctance machines are well known and described in, for example. "Characteristics, design and application of motors and switching reluctance motor devices" by Stephenson and Blake, PCUM '93 Nürnberg, 21-24 June 1993, incorporated herein by reference. This document describes in some detail the characteristics of the switched reluctance machine which together produce REF: 32237 cyclically variable inductance characteristic of the phase coils. Figure 1 shows the main components of a typical switched reluctance motor system. The DC (Direct Current) input power supply 11 may be a battery or a rectified and filtered AC (Alternating Current) supply and may be of a fixed or variable magnitude. In some known motor devices, the power supply 11 includes a resonant circuit which produces a DC voltage which rapidly varies between zero and a predetermined value to allow switching of the zero voltage of the power switches. The DC voltage provided by the power supply 11 is switched through the phase coils 16 of the motor 12 by means of a power converter 13 under the control of the electronic control unit 14. The commutation must be correctly synchronized to the rotation angle of the rotor for the proper operation of the motor device. Typically, a rotor position detector 15 is used to supply signals that indicate the angular position of the rotor. The output of the rotor position detector 15 can also be used to generate a speed feedback signal. The position detector of the rotor 15 can take many forms, for example it can take the form of physical components, as shown schematically in Figure 1. In some systems, the rotor position detector 15 can comprise a transducer of the rotor position that provides output signals that change the state each time the rotor rotates to a position when a different switching arrangement of the devices in the power converter is required 13. In other systems, the position detector can be a programming algorithm which calculates or estimates the position of other verified parameters of the motor system. These systems are often called "sensorless position detector systems" since they do not use a physical transducer associated with the rotor that measures the position. As is known in the art, many different methods have been proposed on the question of a reliable sensorless system. Some of those methods are described later. The energization of the phase coils in a switched reluctance machine depends on the detection of the angular position of the rotor in relation to the stator. This can be explained by referring to Figures 2 and 3, which illustrate the switching of a reluctance machine that operates as a motor. Figure 2 generally shows a rotor 24 with a pole of the rotor 20 approaching a stator pole 21 of a stator 25 according to arrow 22. As illustrated in Figures 2 and 3, a portion 23 of a full-phase coil 16 is wound around the stator pole 21. When the portion 23 of the phase coil 16 around the stator pole 21 is energized, a force will be exerted on the rotor, which tends to pull the pole of the rotor 20 towards the alignment with the stator pole 21. Figure 3 shows generally the typical switching circuit in the energy converter 13 which controls the energization of the phase coil 16, including the portion 23 around of the stator pole 21. When switches 31 and 32 are closed, the phase coil is coupled to the DC power source and energized. Many other configurations of the geometry of the alignment, coil topology and switching circuit are known in the art: some of those are discussed in the documents of Stephenson and Blake cited above. When the phase coil of a switched reluctance machine is energized in the manner described above, the magnetic field established by the flow in the magnetic circuit gives rise to the circumferential forces which, as described, act to pull the poles of the rotor in line with the stator poles. In general, the phase coil is energized to effect rotation of the rotor as follows. In a first angular position of the rotor (called the "ignition angle", TENDERED), the controller 14 provides switching signals to turn on both switching devices 31 and 32. When the switching devices 31 and 32 are turned on, the coil phase is coupled to the DC collective conductor, causing an increase in the magnetic flux that must be established in the machine. The magnetic flux produces a magnetic field in the air space, which acts on the poles of the rotor to produce the verification torque. The magnetic flux in the machine is supported by the magnetomotive force (mmf) which is provided by a current flowing from the DC supply through the switches 31 and 32 and the phase coil 16. Only the current feedback is used. and the magnitude of the phase current is controlled by cutting the current by rapid switching of one or both switching devices 31 and / or 32 on and off. Figure 4 (a) shows a typical current waveform in the cut operation mode, where the current is cut between two fixed levels. In the verification operation, the TENDERED ignition angle is often chosen to be the position of the rotor where the center line of an internal pole space on the rotor is aligned with the center line of a stator pole, but may be some other angle. In many systems, the phase coil remains connected to the collective DC conductor (or intermittently connected if the cut is used) until the rotor rotates so that it reaches what is known as the "free rotation angle", TE_ ,. When the rotor reaches an angular position corresponding to the free rotation angle (e.g., the position shown in Figure 2) one of the switches, for example 31, is turned off. Accordingly, the current flowing through the phase coil will continue to flow, but will now flow only through one of the switches (in this example 32) and only through one of the diodes 33/34 (in this example 3. 4) . During the period of free rotation, the voltage drop across the phase coil is small, and the flux remains substantially constant. The circuit remains in this condition of free rotation until the rotor rotates to an angular position known as the "shut-off angle", STOPPED, (for example when the center line of the rotor angle is aligned with that of the stator pole). When the rotor reaches the off angle, both switches 31 and 32 are turned off and the current in the phase coil 23 starts to flow through the diodes 33 and 34. The diodes 33 and 34 then apply the DC voltage of the conductor collective of DC in the opposite direction, causing the magnetic flux in the machine (and therefore the phase current) to decrease. Other switching angles and other current control regimes are known in the art.

When the machine speed rises, there is less time for the current to rise to the cut or suppression level, and the normal device is normally operating in a "single pulse" mode of operation. In this mode, the ignition, free rotation, and off angles are chosen as a function of, for example, the speed and load torque. Some systems do not use an angular period of free rotation, that is, switches 31 and 32 are turned on and off simultaneously. Figure 4 (b) shows a typical single pulse current waveform where the free rotation angle is zero. It is well known that the values of the ignition, free rotation and shutdown angles can be predetermined and stored in some suitable format to be recovered by the control system when required, or they can be calculated or reduced in real time. It should be understood that sensorless position sensing systems must be able to provide rotor position signals in both cut or suppress operation modes and a single pulse if it is desired to obtain the full capacity of the switched reluctance machine. Although many systems without sensors have been proposed, they tend to be limited to any one mode of operation or impose severe restrictions on the operation of the system. A proposed method using diagnostic pulses in the busy phases is described in "A novel sensorless position sensor for SR engine devices" by Mvungi et al. Proc PEVD Conf, IEE Publication No. 324, London, July 17-19, 1990, pp 249-252. Typically, this method is successful in the cutoff or suppression mode, where the rise and fall times of the current are relatively short compared to the total excitation cycle. The document states that a different method is required for high-speed operation (ie a single impulse). A method for high speed operation is exemplified by EP-A-0573198 (Ray), which describes a flow and current measurement method that leads to predictions of rotor position. Many other sensorless position detection systems have been reviewed and cataloged in "Sensorless methods for determining the position of the rotor of switched reluctance motors", Ray et al. Proc Conference EPE'93, Brighton, EU, 13-16 September 93, Vol 6, pp 7 - 13, and it was concluded in that articles that none of these methods was entirely satisfactory for operation on both operating ranges. According to the invention there is provided a method for determining the position of the rotor in a polyphase switched reluctance machine comprising a rotor, a stator and two or more phases of energizable coils, the method comprising: injecting a first diagnostic pulse of default flow link in one of the phase coils; injecting a second predetermined flow-link diagnostic pulse into another of the phase coils substantially simultaneously with the injection of the first diagnostic pulse; determining the first possible positions of the rotor from a detected characteristic of the first pulse; determining the second possible positions of the rotor from a detected characteristic of the second pulse; and resolving the position of the rotor ambiguously by means of a comparison of the first and second possible positions of the rotor. The invention advantageously exploits the realization that the ambiguity of a point on a phase inductance cycle is closer to one end of the cycle than the other that can be resolved by comparing a pair of such pulses in separate phases. There will only be one point in the cycle where the detected characteristics of both impulses coincide. Preferably, the detected characteristic is the current in the coil. The possible values of the position of the rotor can be stored in a look-up table. Due to the symmetric relationship that exists between the modes of operation of verification and generation in a machine with symmetry in its magnetic characteristics, only a set of values needs to be stored to cover both modes of operation. The resolution of the ambiguity is carried out conveniently by comparing the indicated rotor positions and choosing one commonly indicated by the detected characteristics. The invention can be used to start a machine or when the rotor position is lost during operation. In the latter case it is preferable to allow the decay of the currents or flows to substantially zero to avoid erroneous position calculations. The invention also extends to a switched reluctance machine comprising a switched reluctance machine comprising a rotor, a stator and a plurality of phases of energizable coils, operable switching means for energizing the phases, position detecting means for deriving a position of the rotor relative to the stator, the position detecting means comprise: means for injecting a first pulse of default flow link diagnosis in one of the phases; means for injecting a second predetermined flow link diagnostic pulse in another phase substantially simultaneously with the injection of the first diagnostic pulse; and means for determining the first possible positions of the rotor from a characteristic of the first pulse and second possible positions of the rotor from a characteristic of the second pulse; and means for resolving the position ambiguity of the rotor by means of an analysis of the first and second possible rotor positions. Preferably, the machine comprises means for actuating the switching means according to the position of the resolved rotor. The invention can be practiced in numerous, some of which will now be described with reference to the accompanying drawings in which: Figure 1 shows the main components of a connected reluctance connector system; Figure 2 shows a schematic diagram of a rotor pole approaching a stator pole; Figure 3 shows the typical switching circuit in a power converter controlling the energization of the phase coils of the machine of figure 1; Figures 4 (a) and 4 (b) illustrate typical current waveforms of a connected reluctance device operating in the cut modes and a single pulse respectively; Figure 5 is a schematic diagram of a connected reluctance system embodying the invention; Figure 6 shows the idealized inductance profiles, the excitation regions and the diagnostic regions of a machine operated in the low speed mode according to the invention; Figure 7 shows the idealized inductance profiles and the possible positions of the reference angle for a machine operated in a high speed mode according to the invention; Figure 8 shows the positions determined during the start-up according to the invention; and Figure 9 is a flow diagram according to the operation of the invention. Referring to Figure 5, a switched reluctance machine system comprises a reluctance machine 42 having a rotor 44 mounted to rotate in a stator 46. The stator has two or more (in this case three) phase 48 coils which they are energizable as referred to above. A conventional positioning arrangement 50 is connected to each phase coil. The connection of only one of the coils to the placement arrangement represented schematically is shown in Figure 5 for the purpose of being clear. The placement arrangement controls the application of the d.c. supply voltage. (direct current) V of a supply 52.

The placement arrangement is controlled by the controller 54 comprising a specific application integrated circuit (ASIC) that is programmed to receive recurring information (i, i ') from each of a pair of coils 48, each by means of a current detected device 56, such as the Hall effect device. The ASIC is notationally illustrated as being connected with a digital analog converter 58 and a look-up table memory 60. In practice, the system can use a multiplexed A / D channel of the two current signals or can use two channels Dedicated, for each current transducer. Such systems per se are known in the art. The current transducers can also conveniently supply signals useful for other current verification functions in the system. The A / D converter 58 is arranged to digit the signals representative of the current values detected by the device 56. The look-up table is accessed by the ASIC to convert the values of the detected current to rotor-like values. The values of the rotor angle for a given current are machine specific, but should be common to the currents detected in both phases on the assumption that the phase arrangements are substantially similar. However, separate look-up tables can be used for each phase where the characteristics of the phase differ to a greater degree. The ASIC is programmed to operate at a low speed (cut off) control regime and a high speed (single pulse) control regime as discussed above and a starting method as will be described below. It will be appreciated that the control function of the ASIC is based on programming systems programmed at this point. In this way, the operation will be partially described by means of the flow chart of Figure 9 (discussed below). When the machine is operating at a low speed, the position of the rotor can be determined by injecting a flow link diagnostic pulse of a predetermined amount into an idle (idle) phase coil. The flow link is an integral way of the time of the electromotive force (emf) applied to the coil, which is given by:? =. (V-iR) dt (1) in which? is the flow link of the coil, V is the effective supply voltage (minus any voltage drop in the 50 power converter), i is the coil current and R is the coil resistance. The current is detected by the current sensing device 56 in each phase coil according to the injected flow link pulses. The integration of (V-iR) can be carried out in the ASIC according to the known methods. As a result, a diagnostic pulse occurs: applying the supply voltage 52; verifying the increasing value of the integral; and removing the voltage when the desired value of the flow link is reached. Knowing the values of the flow link and the current, the position of the rotor in table 60 can be consulted to give the value of the rotor angle corresponding to these values. It should be noted that, if the resistance of coil R is small, the term iR in the equation can be ignored for practical purposes. The low speed mode uses a method to inject diagnostic pulses into an inactive phase. When the flow link pulse reaches the predetermined value, the current is recorded and the phase is turned off. From a current table against the rotor angle for this fixed flow, the position can be read. When the flow link has decayed to zero, a subsequent pulse can be initiated and the process repeated. The repetition speed of the impulses in choice for the system designer: the pulses can be injected at a fixed frequency or a new impulse can be started as soon as the previous measurement is completed and the circuit is ready to start a new division . In general terms, the impulse of the flow link has a peak value of the order of 5% to 10% of the peak flow link of the machine. Particularly low values will generally give an increase in inaccuracy due to measurement noise. Particularly high values will generally give an increase in acoustic noise and / or reduced output due to the negative voltage that is being generated. In addition, the higher the impulse, the longer it takes to reach the peak value and the less certainty there will be in the calculated position. The peak flow link value has to be chosen to suit the circumstances.

Note that it is also possible to use a fixed current pulse reading and read the flow link associated with it to read the position of a position / flow link table. For the verification operation, the pulses are placed in the drop region of the inductance. For the generation operation, the pulses are placed in the region of increased inductance. Provided that the profile of the inductance of the machine is symmetrical, only one set of position vs current data needs to be stored since a simple reflection about the maximum or minimum inductance angle will give the correct position for each mode. The system is shown schematically in Figure 6, where LA, LB and LC denote the idealized inductance profiles in the 3-phase machine, Exc A, Exc B and Exc C denote the excitation angles for the verification operation, the D regions denote the rotor angles from which the phases can be diagnosed. There are alternative methods to calculate the flow link. The integral given in equation (1) correctly allows for voltage drops across the switches and for voltage drops across the coil resistance. However, this involves detecting voltage across each phase coil. In many applications, it is possible to simply integrate the DC link voltage, controlling the integrator by knowing if the switches are on, rotating freely or off. Although less accurate than the method in equation (1), it reduces the amount of physical components required, since only one voltage sensor is necessary. Preferably, the peak of the flow link pulse should occur in the region D indicated in Figure 6 for any speed of the machine. Of course, at rest the duration of the impulse is not a concern. However, it has to be taken into account the fact that, when the speed is raised, the same angle indicated by the D region will be covered in less time. Thus, if the peak value of the diagnostic pulse does not vary with the speed of the machine, it must be large enough to establish the desired level of the flow link, but short enough to be useful at higher machine speeds. Preferably, there is sufficient time to inject two or three pulses in region D, so that several position measurements are possible in each diagnostic period. It can be used in the scrutiny of a set of measurements to reduce the effect of any significant inaccuracies described below. Multiple measurements in the same diagnostic period also provide updated position information which can be used for beneficial effects when, for example, changes in machine speed. The high-speed mode, in contrast, interrogates an active phase and takes phase data only once per inductance cycle. An angular reference point is predetermined and the flow of the flow link at this point is measured. Any error between the measured and expected flow link is used to derive a position error and consequently a revised position estimate. The arrangement is shown schematically in Figure 7, where LA, LB and LC denote the idealized inductance profiles of a 3-phase machine and Ref A, Ref B and Ref C denote the reference angles for the three phases for the operation of check. As for the low speed mode, the verification and generation operation can be carried out by exploiting the symmetry of the inductance profile. In any of the low speed or high speed modes, position estimation can be used as the basis of a speed and / or acceleration calculation. Both low speed and high speed modes require at least a rough initial knowledge of the rotor position to operate successfully. In low speed mode, the appropriate region for the injection of the diagnostic pulses must be found: in the high-speed mode, the flow and current link measurements should be taken sufficiently close to the reference angle so that the position error is small. However, when the machine is at rest, or the speed has decreased, or if a transient disturbance in the load or control system causes loss of position data, there is no approximate knowledge of the position and it is extremely unlikely that the system resynchronizes itself accordingly. Prior art control methods of low and high speeds have several drawbacks, since they do not provide a reliable method to start the machine or recover the operation if a transient event occurs causing the loss of position detection. The invention provides techniques to overcome those deficiencies. If there is no knowledge of the position of the rotor and it is required to start (or restart) the excitation of the coil, diagnostic pulses can be injected in a phase, for example Phase A as shown in Figure 8. However, There is ambiguity due to the current measurement and the subsequent calculation. Let's say that A2 would also correspond to the position of Al. If, however, a simultaneous measurement of phase B is made, there are positions Bl and B2. Since there must be a unique position at any angle, only the two points that are equal (A2 and Bl) can be corrected and therefore the position of the rotor can be determined. This is based on the measurements that are being taken practically simultaneously, so it is important that any diagnostic pulse that is used reaches its predetermined value substantially at the same time. This condition is satisfied when flow link pulses are used since, when the connector device is operated from a collective conductor of substantially constant DC, the time it takes the coil to reach a given flow link is effectively determined by the voltage of the collective driver, provided that the phases are identical in number of laps and resistance. For this reason, the systems of the prior art using fixed current value pulses are not suitable, even when they started simultaneously, they would reach their predetermined value at different times and consequently introduce errors in the calculation of the position. The method to determine the position of the rotor is illustrated in the flow diagram of Figure 9 which illustrates part of the programs and programming systems in the ASIC. For the method to work satisfactorily, the flow link and phases must initially be zero and any residual flow link will cause the integration process to start from the wrong starting point. The program determines in block 70 if the machine is recently operating having access to the time since the last switch was operated. If it is determined that a sufficient period of time has elapsed since the last switching activity so that the coil currents (following flow link) is zero, the program comes from block 72. If the elapsed time is insufficient, there is a delay of block 73 before block 72 to allow time for the currents and the flow of the coils to have decayed. In block 72, the diagnostic flow link pulses are injected into two of the machine phases by operating the switching arrangement appropriately. The predetermined flow link pulses are produced by actuating the switching arrangement, so that the DC supply is applied to the two coils simultaneously and for the same duration. According to the same block 72, the currents of the two phase coils are detected by the current detector devices and, at the end of the diagnostic pulse, converted by the A / D converters 58 and the controller 54 into a pair of digital values. The look-up table 60 of the controller 54 is driven by the ASIC in block 74 with the value of the detected current to provide the two angular positions on each of the phase inductance sites for the verified phases at which the detected value of the current will occur for the value of the given flow link of the impulse. To save storage space, it is possible to store only one value of the angle for each current value and use the symmetry of the magnetic characteristics to deduce the other angles. As mentioned above, there is a unique position corresponding to the combination of pairs of current values for each of the phases. Thus, although an ambiguity for which a pair of angular positions is denoted by the value of the single-phase current, the combination of current values for a pair of phases will have only one rotor angle which matches substantial, since the other values of the current will denote disparate positions on the respective rotor position cycles. In this way, the ASIC is programmed to make a comparison of the possible angles of the rotor and to choose the angle that is common to both current readings. Some error in the readings may be presented. In this way, the ASIC can be programmed in any of one of the ways to accept angles of which are close to each other within a tolerance level. In block 75 the ASIC is programmed to determine the appropriate phase to energize based on the calculated rotor angle. The determination of which energized phase is based on the position of the rotor poles which are in the best torsion-producing position (for the desired direction of displacement) for one phase in preference to the others. Based on the decision, the appropriate switches are operated to apply the voltage supply V to that phase coil for its excitation in block 76. The ASIC then cedes normal control of the motor over another programming routine in the ASIC in the block 77. By means of the invention, the position of the rotor is now established. The position of the rotor can be established when the rotor is stationary or running so that a sufficient time interval has elapsed for the current and the flux in the coils to decay so that the appropriate readings of the phases are taken accurately. verified. Sensorless position detection systems generally have to operate in electrically noisy environments near power switching devices, and this often leads to corruption of flow and current link measurements, which leads to data calculations of pure positions. To improve the robustness of the system, a method has been developed to verify the validity of the calculated position data. Each time a new position is calculated, the position, time and speed values can be stored. Using the last n stored values, a target position can be extrapolated compared to the newly calculated one. If the newly calculated and estimated values do not match a predetermined amount, the error control is increased and the estimated value is used instead of the calculated one: if they do not match, any error count decreases and the calculated value is used. At rest, the number of interactions may be limited, but they can still be used to verify the position of the rotor. Particularly under normal operation during successive measurement cycles, a reliability table of the position data is created. If the error count exceeds a certain representative value of, say 5 consecutive calculations without concordance, the control system may decide that it has lost synchronization with the actual position of the rotor and prevents (or interrupts) the excitation of the machine before more serious events occur. The storage and scanning of the values can be effected by any convenient means, but typically by digital storage in memory locations. It has been found that using n = 8 gives a good compromise between the stability of the system and the storage space. It will be noted that this point of the method of the invention is carrout during a repetition of the blocks 72 and 74 by a repetition according to block 78. As indicated above, this interactive process can be implemented in the diagnosis of the position of the rotor in the machine running as well as when the rotor position is established or reset. In this way, the series of n measurements can be used to construct the bases to verify the reliability of the (n + l) th measurement before applying the excitation to the phases. These measurements can be taken from successive pulses. It should be understood that this method of the invention can be used when a position data loss situation occurs and two phases are available for diagnosis. When starting at zero speed or when it is operating initially, those conditions are immediately satisf If a transient disturbance occurs in the load or control system that produces a sudden loss of position data, the conditions can be satisfby removing the excitation of all (or at least two of) the phases and allowing sufficient time for the currents of the phase decay to zero. For example, the peak flow link can be estimated and given time to decay to the estimated zero or the current to be verifby means of current transducers. The two phases can then be diagnosed as described to produce the required position data and the appropriate excitation applto the two phases to produce the required torque to the required direction. Alternatively, if a data reliability verification system is being used while the machine is running, as described above, then the start can be overridden and only a couple of measurements made on the two phases as described. Alternatively, if the coil is put into the free rotation condition after the required flow link is reached, successive measurements of the extended pulse can be taken.

The above description has been based on a system which uses a look-up table of rotor angles for the current values. This is convenient in a digital implementation of the control system. However, it is also possible to use analytical methods to determine the position by inserting the current value measured into a formula that describes the relationship between the current and the angle of the rotor to the flow link that is being used for diagnosis. Such a method may be preferable for the look-up table if the digital storage space is limited and the use of a small table results in highly unacceptable quantization errors. Any of the analytical expressions could be used, for example the Frohlich relation described by Miller et al in "Non-linear theory of the switched reluctance engine for a computer-assisted rapid design" in Proc IEE Pt B. Vol 137, No 6. November 1990, pp. 337-347 or the use of calibration curves according to what was described by Pirón in "The application of magnetic calibration curves to linear motion solenoid actuators and rotating double projection reluctance machines" in ICEM ' 98 International Conference on Electrical Machines, September 2-4, 1998, Istanbul, Turkey, Vol 3, pp 1674 - 1679.

It should be understood that, although the above examples have been described in relation to a three phase machine, the invention can be applied to any polyphase switched reluctance machine with any number of poles. Similarly, the invention could be applied to a linear machine where the moving part (often referred to as a "rotor") moves linearly. Thus, one skilled in the art will appreciate that variations to the described arrangements are possible without departing from the invention. Consequently, the previous description of the different modalities was made by way of example not for the purpose of limiting. The present invention is intended to be limited only by the spirit of the following claims.

It is noted that in relation to this date, the best method known to the applicant to carry out the aforementioned invention, is that which is clear from the present description of the invention.

Claims (19)

CLAIMS Having described the invention as above, the content of the following claims is claimed as property:

1. A method for determining the position of the rotor in a polyphase switched reluctance machine comprising a rotor, a stator and two or more phases of energizable coils, the method is characterized in that it comprises: injecting a first predetermined flow link diagnostic pulse in one of the phase coils; injecting a second predetermined flow-link diagnostic pulse into another of the phase coils substantially simultaneously with the injection of the first diagnostic pulse; determining the first possible positions of the rotor from a detected characteristic of the first pulse; determining the second possible positions of the rotor from a desired characteristic of the second pulse; and solving the position ambiguity of the rotor by comparing the first and second possible rotor positions.

2. The method according to claim 1, characterized in that the detected characteristic of the first and second pulses is the current in the coil.

The method according to claim 1 or 2, characterized in that it includes storing in a look-up table pairs of possible rotor position values for each of a range of characteristic values of the first and second pulses.

The method according to claim 3, characterized in that the look-up table stores a single set of rotor position values for values of the detected characteristic for verification and generation.

5. The method according to claim 1 or 2, characterized in that the look-up table stores a single value of the position of the rotor for each of a range of values of the characteristics of the first and second pulses, the method includes deriving the other possible values by calculating the symmetry of the magnetic characteristics of the machine.

The method according to claim 5, characterized in that the look-up table stores a single set of rotor position values for values of the detected characteristic for verification or generation.

7. The method according to any of claims 1 to 6, characterized in that resolving the ambiguity includes comparing the possible positions of the rotor indicated by the characteristics of the first and second pulses and selecting the position of the rotor commonly indicated by the detected characteristics.

The method according to any of claims 1 to 7, characterized in that the first and second diagnostic pulses are injected by actuating the switching means.

The method according to any of claims 1 to 8, characterized in that the successive pairs of diagnostic pulses are injected at a fixed frequency.

The method according to any of claims 1 to 8, characterized in that the successive pairs of diagnostic pulses are injected at a variable frequency.

The method according to any of claims 1 to 10, characterized in that the first and second diagnostic pulses are injected only when the flow in the respective coils is substantially zero.

The method according to claim 11, characterized in that it includes delaying the injection of the pulses to allow the decay of the flows substantially to zero.

The method according to any of claims 1 to 12, characterized in that it includes verifying the position of the rotor by comparing a first determination with at least a second determination of the position of the rotor.

The method according to claim 13, characterized in that the machine is energized or de-energized depending on the verification of the position of the rotor.

15. The method according to any of claims 1 to 14, characterized in that the determination is made while the rotor is stationary.

16. The method according to any of claims 1 to 14, characterized in that the determination is made while the rotor is in motion.

17. The method according to claim 16, characterized in that it includes de-energizing the coils and, subsequently, delaying the injection of the first and second pulses to allow the coil currents to fall to zero.

18. A switched reluctance motor device comprising a machine having a rotor, a stator and a plurality of phases of energizable coils, operable switching means for energizing the phases, position detecting means for deriving a position of the rotor relative to the stator and means for actuating the switching means according to the position of the rotor, the position detecting means are characterized in that they comprise: means for injecting a first predetermined flow link diagnostic pulse in one of the phases; means for injecting a second predetermined flow link diagnostic pulse in another phase substantially simultaneous with the injection of the first diagnostic pulse; means for determining the first possible positions of the rotor from the first pulse and the second possible position of the rotor from a characteristic of the second pulse; and means for resolving the ambiguity of the rotor position by means of an analysis of the first and second possible rotor positions.

19. The switched reluctance motor device according to claim 18, characterized in that it includes at least one current verifier arranged to detect the current in a corresponding phase coil and in which the characteristic of the first and second pulses is the current.